Objectives: One of the most prevalent complications of orthotopic liver transplant is primary graft dysfunction. Recent studies have shown the preconditioning effect of remifentanil on animal livers but not human livers. Here, we compared the preconditioning effects of remifentanil and fentanyl in orthotopic liver transplant in human patients.
Materials and Methods: In this double-blind clinical trial, 100 patients who underwent liver transplant from deceased donors were randomly allocated into 2 groups. Patients in the remifentanil group received remifentanil infusion, and those in the fentanyl group received fentanyl infusion during maintenance of anesthesia. Serum aminotransferase levels, prothrombin time (international normalized ratio), partial thrombin time, arterial blood gas levels, and renal function tests were evaluated over 7 days posttransplant. Intensive care unit stay and hospitalization were also recorded.
Results: The median peak alanine aminotransferase level during 7 days after transplant was 2100 U/L (interquartile range, 1230-3220) in the remifentanil group and 3815 U/L (interquartile range, 2385-5675) in the fentanyl group (P = .048). Metabolic acidosis, renal state, prothrombin time (international normalized ratio), and partial thrombin time were similar in both groups (P > .05). Durations of stay in the intensive care unit and hospital were not significantly different between the 2 groups (P = .75 and P = .23, respectively). Overall, the clinical outcomes were similar in the remifentanil and fentanyl groups (P > .05).
Conclusions: We found that remifentanil and fentanyl were not different with regard to their preconditioning effects and graft protection in orthotopic liver transplant recipients.
Key words : Intravenous opioid anesthetics, Ischemia-reperfusion injury, Ischemic preconditioning
Early allograft dysfunction is one of the most important complications following orthotopic liver transplant (OLT).1-10 Ischemia-reperfusion injury (IRI) is the main cause of early allograft dysfunction during OLT.6,11-13 Kupffer cells initiate this process, and then the cascade of mediators and cytokines is activated. Finally, apoptosis is the major mechanism for cell injury in IRI.3,6,9,11,13-15 There are several methods that reduce IRI,3,7,12,16-18 among which ischemic preconditioning is an acceptable method against ischemic injury.2,6,9,15
A preconditioning effect can be induced during surgery through various methods. One common method is anesthetics, such as intravenous anesthetic drugs and volatile anesthetics, especially isoflurane, desflurane, and sevoflurane.3,7,12,13,19-22 Many studies have shown the preconditioning effects of these drugs on grafts during OLT.1-3,5-7,13,23-26 Recent studies on rats have shown the preconditioning effect of remifentanil on liver during major liver operations.6,8,22,27 However, such an effect has not been evaluated in humans.
Our present study was designed to evaluate the preconditioning effects of remifentanil as one of the maintenance anesthetic drugs during liver transplant surgeries and to compare its preconditioning effect with fentanyl. It was hypothesized that remifentanil has a greater preconditioning effect than fentanyl on the grafted human liver.
Materials and Methods
This parallel double-blind clinical trial with balanced randomization was conducted from February to August 2018 on 100 patients undergoing liver transplant in the main Transplantation Center affiliated with the Shiraz University of Medical Sciences. Patients ranged from 18 to 65 years old and had no history of previous graft, hepatopulmonary syndrome, portopulmonary hypertension, renal insufficiency (creatinine > 1.5 mg/L), and allergy to propofol, soy beans, and eggs. Deceased donors with cardiopulmonary arrest or hemodynamic instability with high infusion doses of inotropes (infusion norepinephrine above15 μg/min) and cold ischemia time of more than 12 hours were excluded.
The study was approved by the local Ethics Committee of Shiraz University of Medical Sciences.
Randomization and blinding
After written informed consent was obtained from the participants, an anesthesia nurse, who was blinded to the study protocol, allocated eligible patients into 2 equal groups of fentanyl (n = 50) and remifentanil (n = 50) through a simple randomization technique (www.randomization.org). Both participants and care providers were blinded in this study.
Patients were monitored with electrocardiogram, pulse oximetry, noninvasive blood pressure monitoring, temperature monitoring, and bispectral index (BIS). All patients received 30 to 50 μg/kg midazolam and 3 μg/kg fentanyl as premedications. Anesthesia was then induced via intravenous propofol (1.5-2 mg/kg) and atracurium (0.5 mg/kg) as a muscle relaxant. Based on the anesthesiologist’s decision, patients with severe ascites and the possibility of aspiration and difficult intubation were managed as full stomach patients, which required induction with 2.5 mg/kg propofol and 2 mg/kg succinylcholine. After induction of anesthesia and endotracheal intubation, both groups received 100 to 150 μg/kg/min intravenous propofol as a base of maintenance of anesthesia to achieve BIS goal (range, 40-60). The patients were ventilated at 7 to 10 mL/kg tidal volume and positive end-expiratory pressure of 0 to 5 cmH2O. An arterial line was then inserted through the left radial artery, and a large-bore intravenous catheter (No. 12) was inserted through the right internal jugular vein for hemodynamic monitoring. If the mean arterial pressure was < 60 mm Hg, then 0.05 to 0.15 μg/kg/min norepinephrine infusion was initiated.
An anesthesia nurse not involved in the study process prepared 2 identical 50-mL syringes, labeled as A and B. Syringe A contained fentanyl with dilution 50 μg/mL, and syringe B contained remifentanil with dilution 40 μg/mL. In the fentanyl group, a continuous infusion of syringe A was initiated with doses of 2.5 to 9 mL/h (0.03-0.1 μg/kg/min). In the remifentanil group, syringe B was adjusted to 10 to 40 mL/h (0.1-0.4 μg/kg/min) by infusion pump.
Depth of anesthesia was monitored with BIS, and the BIS value was maintained within 40 to 60 by changing the propofol infusion rate. The goal for blood product transfusion was hematocrit level of lower than 30% for packed cell transfusion. Moreover, according to rotational thromboelastometry, fresh frozen plasma, cryoprecipitate, platelet, and fibrinogen were transfused.
All livers were processed by the transplant surgery team; back-table biopsy was done and sent to a single pathologist to evaluate the percentage of macrovesicular steatosis. The surgery was performed via 2 techniques (caval replacement and piggyback). After cavo-caval and portal vein anastomoses, the liver was reperfused. This was followed by arterial anastomosis and bile duct anastomosis. After the surgery, patients were transferred to the intensive transplant care unit for further postoperative care.
Primary outcomes included changes in aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase, prothrombin time, partial thrombin time, and international normalized ratio from day 1 to 7 posttransplant. These parameters were considered to reflect early graft dysfunction.
Secondary outcomes were blood urea nitrogen (BUN), creatinine, and metabolic acidosis (pH and base excess). These were also evaluated from day 1 to 7 posttransplant. Duration of hospital and intensive care unit stay, ventilator need, and extubation time were assessed as secondary endpoints. A data sheet was used to collect the baseline characteristics of recipients and donors and postoperative data of recipients. Cold and warm ischemia times and the blood products transferred during the operation were also recorded.
Sample size and statistical analyses
To report 200 ± 300 U/L decreases for ALT and AST levels and based on the sample size formula with 5% error (α) and power of 90%, 48 individuals were recommended to be allocated in each group. For our study, we enrolled 50 patients.
The normal distribution of data with n < 35 was assessed with one-sample Kolmogorov-Smirnov test. To compare the quantitative variables between the 2 groups, independent 2-sample t test and Mann-Whitney U test were used for normal and nonnormal data, respectively. Data were reported as means and 95% confidence intervals or medians and interquartile ranges (IQR), based on their fulfillment of the normality assumption. The features were compared by using chi-square test. For variables with low sample size, the Fisher exact test was used instead.
From February to August 2018, 200 patients underwent OLT, 145 of which were eligible to be enrolled in this study. The drug was discontinued in 45 patients due to hemodynamic instability and need for high doses of inotrope infusion. Hence, 100 patients completed the course of the study. Figure 1 illustrates the CONSORT flowchart for the patients.
All 100 recipients were similar with regard to baseline demographic characteristics, except for age. Likewise, donors were similar with regard to baseline data (Table 1). Evaluations of macrovesicular steatosis percentage in livers showed both groups to be similar (macrovesicular steatosis range was 0% to 60%). Patients were divided into 2 groups: those with 0% to 30% and those with 30% to 60% macrovesicular steatosis. When we compared the remifentanil and fentanyl groups according to the macrovesicular steatosis percentages, no significant difference was detected (P > .05).
The cold and warm ischemia times were similar in both groups. Propofol doses were changed to achieve the goal of BIS range of 40 to 60. Hence, we compared doses of propofol between the 2 treatment groups to investigate whether this drug had any random effects. The comparisons revealed no significant difference with regard to propofol doses between the 2 groups. Coagulopathies, checked by rotational thromboelastometry, and the need for transfusion in both groups were similar (P > .05). The median dose of norepinephrine, which was infused to achieve hemodynamic stability, was not significantly different between the 2 groups (Table 2).
For comparison, the effect of age was adjusted due to the baseline differences. The median peak AST was similar in both groups (P > .05), but the median peak ALT had a statistically significant reduction in the remifentanil group (P > .048) (remifentanil group = 2100 U/L [IQR, 1230-3220 U/L]; fentanyl group = 3815 U/L [IQR, 2385-5675 U/L]). Although the levels of both AST and ALT were lower in the remifentanil group after 7 days, no significant difference was shown between the 2 groups (P > .05). Other features, including international normalized ratio and direct and total bilirubin, were reduced over 7 days posttransplant; however, no differences were observed between the 2 groups. Time had no effect on changes in the quantity of alkaline phosphatase after 7 days (P > .05). The secondary endpoints, including metabolic acidosis (pH/base excess), improved over 7 days posttransplant, although levels were similar in both groups. Baseline BUN/creatinine showed a statistically significant difference, but these differences were not clinically significant. Blood urea nitrogen and creatinine increased on days 3 and 4 but later reduced to baseline levels. Overall, renal function did not change within 7 days posttransplant (Table 3).
There was only 1 case of mortality in the fentanyl group. Because of this low rate, we did not compare the 2 groups with regard to the mortality rate. The 2 groups were similar in terms of the duration of hospitalization and intensive care unit length of stay, ventilator need, and extubation time.
This randomized clinical trial assessed the preconditioning effect of remifentanil on the transplanted liver in humans. No differences were found in clinical outcomes between the remifentanil and fentanyl groups. This can be due to the protective effect of propofol in the liver of both groups or the similarity in the preconditioning effects of these 2 opioids.
A number of studies have compared the preconditioning effect of hypnotic drugs like volatiles and propofol. In these studies, opioids (remifentanil, fentanyl) were used as analgesic drugs during the operation.3,5,7-9,12,23-25 Here, we compared the preconditioning effect of remifentanil and fentanyl, with propofol used as a baseline anesthetic drug.
Similar to other studies, the present investigation assessed the aminotransferase levels, coagulopathies, renal function, hospitalization and intensive care unit length of stay, ventilator need duration, and extubation time during 7 days posttransplant. We also evaluated the metabolic acidosis states in our patients to obtain the best clinical outcome. Peak ALT was significantly (P = .048) reduced in the remifentanil group (median peak ALT of 2100 U/L [IQR, 1230-3220 U/L] in the remifentanil group); however, this decrease was not clinically significant. Although all biochemical data were lower in the remifentanil group, the 2 groups were not significantly different in this regard (P > .05) (Table 3). All baseline characteristics, macrovesicular steatosis percentages, hemodynamic status, and intraoperative factors (warm and cold ischemia times, coagulopathy state) were similar in both groups (Tables 1 and 2). Age was adjusted for all comparisons and had no influence on the outcomes. The renal state in recipients was compared with BUN and creatinine levels. We observed no baseline clinical differences in levels of these markers; thus, BUN and creatinine did not interfere with outcomes in our investigation. Although we observed a peak in renal markers on days 3 and 4, these levels later reduced to baseline (Figure 2) and no changes were noted with regard to renal status. This reversible peak was predictable and considered due to clamping of the inferior vena cava and portal vein and bleeding during transplant.
Several studies have compared propofol with volatiles, especially sevoflurane. These studies reported no differences in their preconditioning effects on the liver.7,8,23 Some studies have preferred sevoflurane,12,19 whereas others have preferred propofol.24,25,28 Therefore, the present study decided to use total intravenous anesthesia and used propofol infusion with BIS monitoring to achieve the best depth of anesthesia and optimum conditions during the transplant procedure since neuroendocrine changes occurring during light anesthesia might affect IRI. Opioids have only been evaluated in animal studies, and the effects of morphine have been compared intravenously versus intrathecally in rat livers.15 Recent studies have assessed the preconditioning effects of remifentanil on animal organs, including the liver, kidney, and heart in rats and rabbits.10,21,22,26,27,29,30 All of these studies used remifentanil per se or compared it with placebo.
In a study from Yang and associates, infusion of remifentanil was compared versus normal saline (placebo) in 50 rats undergoing liver operations.22 In a study from Zhao and colleagues, remifentanil was infused in 1 group of rats, with the other group not receiving any drug during liver surgical procedures. They showed an antiapoptotic effect of remifentanil on liver cells.10 In addition, remifentanil was shown to attenuate inflammatory responses in rats, as shown by examination of AST and ALT levels, biochemical cytokines, mediators, and liver biopsies in vitro.26 Remifentanil was recently shown to have a good preconditioning effect on the liver and markedly reduced aminotransferase levels. Mechanisms included reduction of free oxygen radicals, deactivation of myeloperoxidase, and malonyl dialdehyde producing nitric oxide; thus, remifentanil attenuates the inflammatory responses and protects the cell from injuries with antiapoptotic activity.10,22,26
The strength of our study was its double-blind randomized clinical trial design following the CONSORT guideline. All surgical procedures were performed by an expert anesthesiologist and the transplant team. Considering the limitations of liver transplant procedures and our exclusion criteria, our sample size seemed to be statistically appropriate; nonetheless, further studies with larger sample sizes are warranted to confirm the present findings.
No serious complications occurred during the study, with only 1 mortality in the fentanyl group.
In this double-blind randomized clinical trial, the final clinical outcomes were more important than the biochemical and histologic outcomes. Here, we found the clinical outcomes to be the same in both treatment groups. This finding is significant because remifentanil is an expensive and not easily accessible drug. Therefore, despite its good preconditioning effect on animals, remifentanil seems not to be superior to fentanyl in humans.
DOI : 10.6002/ect.2019.0014
From the 1Department of Anesthesia, Nemazee Hospital, and the 2Anesthesiology
and Critical Care Research Center, Department of Anesthesia, Nemazee Hospital,
Shiraz, Fars, Iran
Acknowledgements: The authors have no conflicts of interest to declare. The registration number of the study is IRCT20141009019470N70. This work was supported by the Vice-Chancellery of Research and Technology of Shiraz University of Medical Sciences (Grant 95-01-01-11395). This article was based on the thesis written by Sanaz Jowkar, MD, which was submitted to the School of Medicine in partial fulfillment of the requirements for the degree specialty in Anesthesiology. The authors thank the Center for Development of Clinical Research of Nemazee Hospital. Appreciations are also expressed to Dr. Saeede Pour Ahmad from the Research Consultation Center (RCC) of Shiraz University of Medical Sciences for the statistical analyses and Ms. Farzaneh Rasouli for proofreading, editing, and improving the English structure of this manuscript.
Corresponding author: Mohammad Bagher Khosravi, Anesthesiology and Critical Care Research Center, Department of Anesthesia, Nemazee Hospital, Zand Street, Shiraz, Fars, Iran 71937-11351
Phone: +98 7136474270
Table 1. Comparison of Baseline Characteristics of Recipients and Donors
Table 2. Comparison of Intraoperative Parameters Between the 2 Groups
Table 3. Comparison of Postoperative Biochemical Outcomes Between the 2 Groups
Figure 1. CONSORT 2010 Flow Diagram
Figure 2. Pattern of Changes in Mean Creatinine Levels During 7 Days Posttransplant